1% risk per trade

options-george

Guest Author
Messages
484
Likes
94
It seems to be a common risk management for full-time retail traders to set a risk of 1% of their trading account on each trade, thus for someone with a £100k account, they risk £1,000 per trade by setting the stop loss for the trade in the relevant place.

Is there a rationale for setting this to 1%? Why not set it to 1.5%, 2% for example?

Many thanks in advance,
George
 
2% is fine as well, it's just more risk so returns should be larger if profitable but then so will drawdowns.

Ultimately the rational for 1% or 2% is this - if you don't have a way to control your losses the markets will at sometime eat you alive, however good a trader/investor one is.

The streets of the City and Wall St are littered with great traders, sometimes brilliant traders where the risk got them.

Also, studies have shown that anything between about 5%-10% (nearer 10%) the risk-of-ruin of a trader going broke is 1. It might not happen for 10 or more years (might happen next week) but it's going to happen........
 
The question is now this George. Do you think the above applies to you or not? Many think it doesn't and then.........
 
The question is now this George. Do you think the above applies to you or not? Many think it doesn't and then.........

Hi Anley
Thanks for your replies.

My reason for asking is a mathematical one. In the gambling that I have done we tend to bet the % of the bankroll that equates to the estimated advantage on a particular bet. Thus if we have a 1% edge then we bet 1% of the BR, if we have a 2% edge then we bet 2% of the BR and so on. I actually often discounted the % further to allow for more fractional Kelly betting.
In these situations the edges were statistically calculated.

I have been trying to compute edges for the FX trading as well. Clearly the outcome of the markets cannot be modelled with certainty as the outcomes of the spin of a roulette can be. However I have been trying to estimate the edge by looking at the Win% and the RR ratio and calculating the edge in this manner.

Let's assume that a trader wins 45% of his trades, break-evens on 15% and 40%, and achieves a R:R of 1.4:1. his risk is $1000/trade. This would be a strategy that results in say 10 trades per month.

In that case I believe the edge would be
(45% x $1400) + (40% x -$1000)
= $630- $400
= $230
(so 10 trades x $230 produces an estimated gain of $2,300/month for the trader)

The EV of $23 constitutes an edge of 23% on the trader's investment. I think the target Win ratio and target R:R is not unrealistic. If the trader has an edge of say 23% it would surprise that he risks only 1% on each trade. His bankroll could grow much quicker if he allowed the risk to be 2-3%, and still stay well within a reasonable ROR level.
 
True re your figures of 23% etc. But what you're not (I think) taking into account is the mental capital, or the psychology of trading.

Keeping your risk small means smaller profits of course but EASY TRADING, ie you put the trade on and if it loses you lose 1%, no big deal. So you put the next trade on and it loses so now you're down 2%, no big deal. This goes on for 5 losses in a row and so you're 5% down, again not that much of a problem given you're expected to make an overall profit over the year.

Now let's do it with 3%, well you know it's 15% loss. Easy to talke about but when it's real money you're trading is now a lot different, it's HARD, maybe you're getting a little bit scared etc, maybe you don't take the next trade because it's looks like a sure loser (or course it won't be if you don't take it).

But you might say, that it's never had 5 losers in a row? Sure, but the data sample is getting bigger which means the runs (plus and minus) will also get bigger so a 5 loss run is going to come, just a matter of time.

Summary: With 1% you won't make as much but you'll be able to take every trade and remain upbeat whatever happens. With 3%+ per trade you could become a nervous wreck and so you defeat yourself rather than the market or strategy.

Not sayint you don't have thick blood but many people think that until they find out their blood was very thin, losses can do that to people......
 
What you plebs talking about!!!! We gamblers... If it comes down to the last pip I'll trade my wife in for "favours" to hit the big daddy
 
My reason for asking is a mathematical one. In the gambling that I have done we tend to bet the % of the bankroll that equates to the estimated advantage on a particular bet. Thus if we have a 1% edge then we bet 1% of the BR, if we have a 2% edge then we bet 2% of the BR and so on.

I suggest you review your understanding of Kelly criterion. Here is a good start http://en.wikipedia.org/wiki/Kelly_criterion

When calculating Kelly bet you need to know probability distribution of different outcomes, not the advantage.

In the example you gave the Kelly bet is about 19% and risking 2-3% is a really conservative approach. On other hand if you are going to bet 23% (your advantage) you will overbet for no good reason.
 
There is no rationale. I wouldn't be the slightest bit surprised if it came from this line of thinking

1-We need to risk a % of the account.
2-A small number is safer.
3-1 is the smallest number.
 
It all dependends on your trading system and your own risk tolerance.

If you only trade one or two markets or stocks at a time then 2, 3, 4 or even 5% risk might be ok depending on your strategy.

If you into trading lots of markets at the same then, say 5 or 20 markets, then 1% or 0.5% is more sensible.

Also you dont have to stick to one risk% number.. when your trading is going well you increase your risk% and when things are going badly you can cut it down.
 
I suggest you review your understanding of Kelly criterion. Here is a good start http://en.wikipedia.org/wiki/Kelly_criterion

When calculating Kelly bet you need to know probability distribution of different outcomes, not the advantage.

In the example you gave the Kelly bet is about 19% and risking 2-3% is a really conservative approach. On other hand if you are going to bet 23% (your advantage) you will overbet for no good reason.

Wow! So with 53% win rate and 1.4 reward, Kelly suggests betting 19% of bankroll/account for optimal EV?
 
Wow! So with 53% win rate and 1.4 reward, Kelly suggests betting 19% of bankroll/account for optimal EV?
No, not for optimal EV. Your EV does not depend on bet size, it is still 23%. Kelly optimises bankroll growth. You do want to increase your bankroll as fast as possible, do you? :cheesy:

Why do you surpised? You have almost 50:50 proposition (think about coin flip slightly biased in your favor) and every time you win you get 1.4 of your stake. This is a really good proposition. 19% sounds reasonable.
 
No, not for optimal EV. Your EV does not depend on bet size, it is still 23%. Kelly optimises bankroll growth. You do want to increase your bankroll as fast as possible, do you? :cheesy:

Why do you surpised? You have almost 50:50 proposition (think about coin flip slightly biased in your favor) and every time you win you get 1.4 of your stake. This is a really good proposition. 19% sounds reasonable.

Yeah I slipped up there as EV is same!

Surprises me as it wouldnt take much to have enough losses to do some serious 'damage' to the account?
 
I like it though; downloaded Kelly's paper :)
I read sometime ago a paper which analyses how the risk of ruin depends on the fraction of Kelly bet. Quite interesting as well. Sorry cannot give you a hyperlink.
 
Thank you so much for all the replies and discussion re my initial question - very appreciated. This question had been on my mind on some time so I wanted to get some other people's thoughts on it :)
 
I've got a follow-up question on this.

Assume that normally you risk 1% per trade.
You put on two trades at the same time - the first you are buying USDCHF, and in the second you are shorting GBPUSD - so in both cases you are long USD.

In this situation, would you then allocate less than 1% risk on each trade?
 
I've got a follow-up question on this.

Assume that normally you risk 1% per trade.
You put on two trades at the same time - the first you are buying USDCHF, and in the second you are shorting GBPUSD - so in both cases you are long USD.

In this situation, would you then allocate less than 1% risk on each trade?

If I was trading any two instruments of any kind that were closely correlated, I would choose the better set up (according to whatever criteria I use) and trade that one.

However, if you insist on trading two things that are going to move in step, I would think of what you are actually doing. You risk 1% (or whatever) but on what? We say "per trade" because that what it normally is, but really it isn't that (or isn't necessarily).

Perhaps we should say "exposure to market movements" - if you put 1% each on two correlated instruments, your exposure is really 2%, even though we look at it as two separate entities.

So to answer your question, I would split my 1% so that my total exposure remains at 1%.

By the way, when I started out in forex I did what you're talking about - by accident. :LOL:
 
Top